5.8.1升级至5.9.0文档

版本变更

5.9.0 版本变更

控制台升级

  • helm

  • Basic Edition

  • 如控制台为 allinone 启动，执行以下操作

  1. docker stop rainbond-allinone && docker rm rainbond-allinone
  2. #该命令参数需要和之前启动的rainbond-allinone容器一致
  3. docker run -d -p 7070:7070 -v ~/.ssh:/root/.ssh -v ~/rainbonddata:/app/data \
  4. --name=rainbond-allinone --restart=always \
  5. registry.cn-hangzhou.aliyuncs.com/goodrain/rainbond:v5.9.0-release-allinone

如控制台为通过应用市场安装，进入控制台应用内，选择升级到 5.9.0 版本

数据中心升级

1. 更新 grctl 命令

  1. docker run -it --rm -v /:/rootfs  registry.cn-hangzhou.aliyuncs.com/goodrain/rbd-grctl:v5.9.0-release copy
  2. mv /usr/local/bin/rainbond-grctl /usr/local/bin/grctl && grctl install

注意：如果遇到无法安装 grctl 可以参考grctl 安装文档

2. 执行升级命令

  1. grctl cluster upgrade --new-version=v5.9.0-release

3.更换 operator 镜像并修改挂载

  1. kubectl edit deploy rainbond-operator -n rbd-system

  • 替换镜像为以下

    1. registry.cn-hangzhou.aliyuncs.com/goodrain/rainbond-operator:v5.9.0-release

  • 修改挂载的配置如下，注意 volumeMounts 以及volume 格式对齐

    1. volumeMounts:
    2. - mountPath: /var
    3.   name: vardir
    4. - mountPath: /run
    5.   name: rundir
    7. volumes:
    8. - hostPath:
    9.     path: /run
    10.     type: Directory
    11.   name: rundir
    12. - hostPath:
    13.     path: /var
    14.     type: Directory
    15.   name: vardir

配置文件如下:

  1. # 需要配置 image volumeMounts 以及 volumes 字段的内容
  3. # Please edit the object below. Lines beginning with a '#' will be ignored,
  4. # and an empty file will abort the edit. If an error occurs while saving this file will be
  5. # reopened with the relevant failures.
  6. #
  7. apiVersion: apps/v1
  8. kind: Deployment
  9. metadata:
  10.   annotations:
  11.     deployment.kubernetes.io/revision: "1"
  12.     meta.helm.sh/release-name: rainbond
  13.     meta.helm.sh/release-namespace: rbd-system
  14.   creationTimestamp: "2022-10-10T03:28:19Z"
  15.   generation: 1
  16.   labels:
  17.     app.kubernetes.io/managed-by: Helm
  18.     control-plane: rainbond-operator
  19.     release: rainbond
  20.   name: rainbond-operator
  21.   namespace: rbd-system
  22.   resourceVersion: "926"
  23.   uid: 5fd3bcf3-52e2-4a29-a653-511bb0cd7058
  24. spec:
  25.   progressDeadlineSeconds: 600
  26.   replicas: 1
  27.   revisionHistoryLimit: 10
  28.   selector:
  29.     matchLabels:
  30.       control-plane: rainbond-operator
  31.   strategy:
  32.     rollingUpdate:
  33.       maxSurge: 25%
  34.       maxUnavailable: 25%
  35.     type: RollingUpdate
  36.   template:
  37.     metadata:
  38.       creationTimestamp: null
  39.       labels:
  40.         control-plane: rainbond-operator
  41.         release: rainbond
  42.     spec:
  43.       containers:
  44.       - args:
  45.         - --leader-elect
  46.         - --zap-log-level=4
  47.         command:
  48.         - /manager
  49.         image: registry.cn-hangzhou.aliyuncs.com/goodrain/rainbond-operator:v5.9.0-release
  50.         imagePullPolicy: IfNotPresent
  51.         livenessProbe:
  52.           failureThreshold: 3
  53.           httpGet:
  54.             path: /healthz
  55.             port: 8081
  56.             scheme: HTTP
  57.           initialDelaySeconds: 15
  58.           periodSeconds: 20
  59.           successThreshold: 1
  60.           timeoutSeconds: 1
  61.         name: rainbond-operator
  62.         readinessProbe:
  63.           failureThreshold: 3
  64.           httpGet:
  65.             path: /readyz
  66.             port: 8081
  67.             scheme: HTTP
  68.           initialDelaySeconds: 5
  69.           periodSeconds: 10
  70.           successThreshold: 1
  71.           timeoutSeconds: 1
  72.         resources:
  73.           requests:
  74.             cpu: 100m
  75.             memory: 64Mi
  76.         securityContext:
  77.           allowPrivilegeEscalation: false
  78.         terminationMessagePath: /dev/termination-log
  79.         terminationMessagePolicy: File
  80.         volumeMounts:
  81.         - mountPath: /var
  82.           name: vardir
  83.         - mountPath: /run
  84.           name: rundir
  85.       dnsPolicy: ClusterFirst
  86.       restartPolicy: Always
  87.       schedulerName: default-scheduler
  88.       securityContext: {}
  89.       serviceAccount: rainbond-operator
  90.       serviceAccountName: rainbond-operator
  91.       terminationGracePeriodSeconds: 10
  92.       volumes:
  93.       - hostPath:
  94.           path: /run
  95.           type: Directory
  96.         name: rundir
  97.       - hostPath:
  98.           path: /var
  99.           type: Directory
  100.         name: vardir
  • 等待 operator 启动完成以后重启 node 以及 chaos
  1. kubectl delete ds rbd-chaos rbd-node -n rbd-system
5.8.1升级至5.9.0文档 - 图1caution

非 helm 安装的 Rainbond，出现以下报错为正常现象，可以忽略

  1. error: update rainbond components: get rbdcomponent rbd-app-ui: rbdcomponents.rainbond.io "rbd-app-ui" not found